COPElAND | 205
a series of lectures on computer design in London, and it was decided that Kilburn would
attend.^44 The lectures ran from December 1946 to February 1947 and were held in a conference
room at the Adelphi Hotel in the Strand.^45
When asked, in later life, where he got his basic knowledge of the computer from, Kilburn
usually said, rather irritably, that he couldn’t remember.^46 In an interview, he commented
vaguely:^47
Between early 1945 and early 1947, in that period, somehow or other I knew what a digital
computer was . . . Where I got this knowledge from I’ve no idea.
However, Kilburn had referred to ‘unpublished work’ by Turing in his first report on his com-
puter research, dated December 1947, and he had used a number of Turing’s technical terms.^48
These included Turing’s ‘universal machine’ and ‘table of instructions’, and also various terms
that, whether or not they originated with Turing, were distinctive of Turing’s approach in the
Adelphi lectures: for example ‘source’, ‘destination’, ‘temporary store’, ‘staticisor’, and ‘dynamici-
sor’. In a subsequent report, written when the computer was working, Kilburn said:^49
I wish to acknowledge my indebtedness to Prof. M. H. A. Newman, and Mr. A. M. Turing for
much helpful discussion of the mathematical requirements of digital computing machines.
Williams summed up Turing’s role like this:^50
[Turing was] instrumental, with Newman, in instructing us in the basic principles of computing
machines, not on the engineering side of course, on the mathematical, and we had very close
collaboration with both of them.
The Adelphi lectures played a key role in the Manchester project, by introducing Kilburn to
the fundamentals of computer design. There is in fact no mystery about where Kilburn got his
basic knowledge of the computer from—Turing taught him.
kilburn's first computer
Kilburn was a good pupil, quickly progressing during the lectures from not knowing the
‘first thing about computers’^51 to the point where he could start designing one himself: in
fact, Kilburn’s initial design for what would eventually be the Manchester computer followed
Turing’s principles closely. Turing, unlike von Neumann and his group, advocated a decen-
tralized computer with no central processing unit (cpu)—no one central place where all the
logical and arithmetical operations were carried out. (The term ‘decentralized’, and its opposite
‘centralized’, were due to Jack Good, for whom Newman had created a special lectureship in
Mathematics and Electronic Computing in his Manchester department.^52 ) Kilburn designed a
decentralized computer very much along the lines that Turing set out in his lectures: it was quite
different from the centralized type of design that von Neumann was proposing.
In Version V of Turing’s ACE design, which formed the subject matter of the first five Adelphi
lectures, there was a collection of different ‘sources’ and ‘destinations’ in place of a central accu-
mulator. The central accumulator is a storage unit that forms the sum of an incoming number
and the number already stored: this sum then replaces the previous contents of the accumulator.
Early centralized computers did all their calculations by means of transferring numbers to and
from the accumulator. In Turing’s decentralized Version V, on the other hand, the arithmetic